Comparing prototype-based and exemplar-based accounts of category learning and attentional allocation

Area of Science:

• Cognitive Psychology
• Categorization Research

Background:

• Exemplar theory, often supported by the 5/4 stimulus set, has been favored over prototype models for explaining categorization.
• The exemplar model uniquely predicts certain qualitative performance aspects in categorization tasks.

Purpose of the Study:

• To re-evaluate the fit of exemplar and prototype models to categorization data using the 5/4 stimulus set.
• To investigate the assumptions made by exemplar models regarding human attention and effort.

Main Methods:

• Conducted two experiments comparing prototype and exemplar models on categorization performance.
• Assessed participants' attentional strategies independently to validate model assumptions.
• Reviewed 30 prior studies using the 5/4 stimulus set.

Main Results:

• Prototype models provided a better fit to participants' performance profiles than exemplar models.
• The exemplar model's explanation for qualitative performance aspects relied on questionable assumptions about attention and effort.
• Independent assessment of attention contradicted the exemplar model's account.

Conclusions:

• Prototype models offer a more parsimonious and accurate explanation of categorization behavior than exemplar models.
• Exemplar models may overstate the role of attention and effort, making biologically implausible assumptions.
• Revisiting the 5/4 stimulus set literature suggests a need to reconsider the dominance of exemplar theory.